Nursing knee pain gets treated as a footwear problem, and the standard advice - wear something flat and flexible, let your foot move naturally - sounds so reasonable that almost nobody checks it. It was checked. In a randomized trial of 164 people with symptomatic knee osteoarthritis, flat flexible shoes were compared against stable supportive ones for six months, and the flat flexible shoes did not win. The stable supportive ones produced less pain.
What the trial found, stated carefully
The study (Paterson and colleagues, Annals of Internal Medicine, 2021) randomized participants with moderate to severe medial knee osteoarthritis to one of two shoe types, worn at least six hours a day for six months. The researchers set out to test whether flat flexible shoes were superior, because earlier laboratory work had shown lower knee loading in that type of shoe. They found no evidence of superiority. On walking pain, the difference favored the stable supportive shoes by 1.1 units (95% CI 0.5 to 1.8), and a similar pattern showed up in secondary measures including knee-related quality of life.
Three honest caveats. The participants had diagnosed knee osteoarthritis, which is not the same population as a 34-year-old nurse whose knees hurt after four shifts in a row. The trial compared shoe categories, not specific models, so it cannot tell you which shoe to buy. And the mechanism is unresolved - the lead researcher suggested cushioning as the likely explanation and said plainly that more research is needed. What the trial does settle is that “flat and flexible is better for knees” is not supported, and the advice you will hear repeated in break rooms has it backwards.
None of us is a clinician. Knee pain that persists, swells, locks or gives way is a reason to see someone, not a reason to shop. What follows is about which shoes the evidence favors, not about treatment.
The criteria that follow from it
- Cushioning, in quantity. It is the mechanism the trial authors themselves proposed. Stack height is the available proxy, and hospital flooring is the surface that makes it matter.
- Stability through the heel and base. A heel that rocks is an ankle doing extra work, and that work gets passed upward. Broad bases and firm heel counters do more here than arch gimmicks.
- Enough structure to stop the shoe folding. “Flexible” was the losing arm of the trial. A shoe you can twist into a spiral is not supporting anything.
- Fit you will actually tolerate for twelve hours. The best-evidenced shoe in the world does nothing from inside your locker.
- Weight, because knees pay for it. Every ounce is lifted thousands of times a shift.
Heel-to-toe drop: why we do not rank by it
Drop - the height difference between heel and forefoot - is the spec most often recommended for knee pain, usually with confidence that the research does not support. The lab studies are small, acute and contradictory. One 2021 study of habitual rearfoot strikers found that larger drops increased the peak knee extension moment while decreasing ankle eversion moment; other work has found lower knee flexion moments in zero-drop shoes during push-off. Different outcome measures, different directions, all in single sessions on a treadmill rather than over a nursing career.
The practical reading is that drop shifts load between your ankle and your knee rather than removing it. If your knees are the problem and your calves are fine, a lower drop is worth trying; if your Achilles and calves are already tight, dropping the heel will move your pain rather than fix it. That is an experiment to run with your own legs, not a specification to buy blind. The shoes below span 6 mm to 10 mm precisely so there is somewhere to go in either direction - and note that HOKA’s measured drops run higher than its listed ones, which the HOKA roundup documents model by model.
What a shoe cannot do for a knee
Three things get sold as knee solutions and are not. Arch supports marketed for knee pain are treating a different joint; they can help if your arches genuinely collapse, which is a specific finding and not a default. Compression sleeves worn on the knee may feel reassuring but are not footwear and are outside what this page can speak to. And cushioned insoles added to an unsupportive shoe do not turn it into a supportive one - if you want to improve a shoe you already own, our insoles roundup is explicit about what an insole can and cannot change.
The variable with the most leverage is probably not the shoe at all: it is how many consecutive twelve-hour shifts you work and whether you ever get off the floor. That is not something we can sell you. It is still true.
Standing still is its own problem
Nursing knees are not a running injury. The pattern is long static standing punctuated by bursts of walking, often on polished concrete, often in the same spot at a workstation for an hour at a time. Static standing loads a joint continuously with no cycle of loading and unloading, which is why the knee can hurt more on a slow shift than a busy one.
Two things follow. Shift your weight deliberately - a foot up on a rail, a step side to side every couple of minutes - because movement is what circulates fluid through the joint. And treat footwear as one input among several: our guides on foot pain from standing all day and compression socks cover the rest of the lower limb, which does not behave independently of the knee.
What to check before buying
- Can you twist the shoe? If it folds easily in your hands, it is the shoe type the trial found inferior.
- Does the heel counter hold firm? Squeeze it. It should resist, not collapse.
- Is the base wide at the heel? Look at the footprint from below; a narrow heel on a tall stack is the wobbly combination.
- Do your old shoes lean? Inward lean means arch collapse, and a stability model is the version of this list to buy.
- Are you replacing worn-out shoes? Compressed foam cushions nothing; some of the best knee results come from replacing a dead pair rather than upgrading a live one.
